EFFECTS OF NITROUS OXIDE ON AUDITORY CORTICAL EVOKED POTENTIALS AND SUBJECTIVE THRESHOLDS

Transcription

1 Br. J. Anaesth. (1988), 61, EFFECTS OF NITROUS OXIDE ON AUDITORY CORTICAL EVOKED POTENTIALS AND SUBJECTIVE THRESHOLDS H. G. HOUSTON, R. J. McCLELLAND AND P. B. C. FENWICK General anaesthetics are known to alter both the spontaneous and evoked electrical activity of the brain. The responsiveness of the auditory group of evoked potentials and ease of stimulation make them particularly suitable for monitoring depth of anaesthesia. Dose related changes have been reported in auditory brainstem and middle latency evoked potentials with a range of anaesthetic agents [1, 2]. The later auditory cortical evoked potentials are affected also by the concentration of anaesthetic. Jarvis and Lader [3] reported a reduction in the amplitude of all components of the auditory cortical evoked potential with increasing doses of nitrous oxide. These findings were confirmed later [4]. The reductions in the amplitude of auditory cortical evoked potentials with nitrous oxide may be secondary to increased threshold of the evoked potential and possibly related to changes in subjective auditory threshold. Burns and colleagues [5] reported increases in subjective auditory thresholds to 500-Hz tones with nitrous oxide. However, Fenwick and colleagues [4] noted that several subjects reported an increase in hearing sensitivity with higher concentrations of nitrous oxide. The present investigation was undertaken to evaluate the origins of the amplitude change in auditory cortical evoked potentials, within the latency interval ms after the stimulus produced by nitrous oxide, and to study the effects of nitrous oxide on both subjective pure tone threshold and the threshold of the cortical evoked potentials. H.G.HOUSTON; R. J. MCCLELLAND; P. B. C. FENWICK* ; Department of Mental Health, Queen's University, Belfast BT9 7BL. Accepted for Publication: February 12, *Present address: St Thomas' Hospital, Lambeth Palace Road, London SE1 7EH. Correspondence to R. J. McC. SUMMARY This study has examined the effects of inhaled nitrous oxide on the N1 and P2 components of the cortical auditory evoked potentials (AEP) in the latency interval ms after the stimulus. The amplitudes, latencies and thresholds of the AEP were measured at a range of end-tidal nitrous oxide concentrations (0%, 10%, 20%, 40%) in 10 subjects with normal hearing. Systematic decreases in amplitude and latency were observed with an increase in threshold. A study of the effect of stimulus intensity on AEP amplitude showed that the amplitude change with nitrous oxide was accounted for largely by systematic increase in evoked potential threshold. Subjective pure tone thresholds were not affected by the concentrations of nitrous oxide used, indicating that the AEP changes were independent of subjective hearing level. SUBJECTS AND METHODS We studied 10 healthy volunteers (five male) aged yr with normal hearing and thresholds equal to or better than 10 db HL (hearing level) across the frequency range. Silver-silver chloride disc electrodes were placed on each mastoid and the vertex and an earth electrode at the forehead. The skin resistance was maintained at 2000 Q. Two channels of 1-s epochs of EEG were recorded: vertex to ipsilateral mastoid and vertex to contralateral mastoid. The responses were amplified by 2 x 10 4, with a signal bandwidth of 1-30 Hz. One ear was tested on each subject. The stimulation procedure consisted of delivering a 20-ms tone burst at 2000 Hz, from an Ampliaid Mark III Stimulator, through a single TDH 39 headphone. The triggering of the stimulator was delayed until 500 ms pre-stimulus EEG had been

2 NITROUS OXIDE AND EVOKED POTENTIALS 607 collected to give a measure of the background EEG, and was of considerable value in establishing the presence of a response. The stimulus rate was once every 3 s. A total of 64 1-s epochs of EEG were averaged using an Apple II microprocessor and the data stored on floppy disc for off-line analysis. Latency and amplitude measurements were made on the grand averages and their two constituent sub-averages, which were filtered digitally (zero phase shift) using a pass band of 1-12 Hz. For each trial, auditory cortical evoked potentials were recorded to a high level stimulus of 70dBHL, 2000-Hz tone burst, followed by measurement of the cortical evoked potential threshold. Finally, the subject's pure tone threshold was determined at 2000 Hz. Nitrous oxide was administered from a Boyle's anaesthetic machine. Each subject was studied with four different mixtures of oxygen and nitrous oxide: 100% oxygen, and 10%, 20% and 40% nitrous oxide in oxygen. The percentage of oxygen delivered was measured using a Critikon Oxychek oxygen analyser and the balance assumed to be nitrous oxide. Ten minutes was allowed for gas equilibration at each concentration. Analysis of variance was applied to each of the variables (amplitude, threshold and latency). Each variable was analysed also using the nonparametric Friedman analysis of variance, in view of the possible heterogenity of variances. These further analyses confirmed the results of the analyses of variance. The project was approved by the Ethics Committee of the Faculty of Medicine, Queen's University, Belfast. RESULTS Initial inspection of the cortical evoked potentials showed that only the first main negative (Nl) and positive (P2) components could be identified readily throughout the test. Amplitude measurements were based on the peak N1/P2 component, and the latencies of Nl and P2 were measured. No differences were observed in recordings from right mastoid and left mastoid for any measurement of amplitude or latency. For each subject the average was taken of two hemisphere measurements of each parameter. The N1/P2 amplitude at 70 db HL decreased with increasing concentrations of nitrous oxide (fig. 1) and was highly significant on analysis of variance (P < ). A normalization procedure was introduced to deal with the wide variations in amplitude of evoked potentials between subjects. The mean corrected amplitudes were obtained by calculating the average amplitude for each subject 5 JV Time (ms) FIG. 1. Mean cortical evoked potentials for 10 subjects. 500

3 608 BRITISH JOURNAL OF ANAESTHESIA OH < Q. -10' FIG. 2. Mean against nitrous 1 h- 50-i Nitrous oxide (%) corrected amplitude for each subject plotted oxide concentration. Slope of regression line = Nitrous oxide CL) FIG. 3. Mean corrected threshold for each subject plotted against nitrous oxide concentration. Slope of regression line = the nitrous oxide concentration and the N1/P2 amplitude (r = -0.78; P < 0.01). In parallel with the amplitude changes, a progressive increase in cortical auditory evoked potential threshold was observed with increasing concentrations of nitrous oxide. Analysis of variance showed this to be highly significant (P < 0.001). There were variations in evoked potential thresholds and mean corrected thresholds for the four conditions were plotted (fig. 3); a significant correlation was found between nitrous oxide concentrations and evoked potential threshold (r = 0.82; P < 0.01). One explanation for the change of amplitude with nitrous oxide, therefore, was an increased evoked potential threshold. The mean shift in threshold in evoked potential between oxygen and 40 % nitrous oxide was 33 db. In an attempt to examine the relationship between threshold increase and amplitude decrease, the effects were investigated directly of stimulus intensity on evoked potential amplitude. Cortical evoked potentials were obtained from volunteers at 70 db HL (the same hearing level used in the nitrous oxide experiment) and at 40 db HL. It was found that reducing the stimulus by 30 db reduced the mean amplitude of the cortical evoked potential by 4.0 nv (table I). This was similar to the mean decrease in amplitude induced by 40% nitrous oxide (4.9 nv). Subjective pure tone thresholds to 2000-Hz tones were measured while the subject was breathing oxygen and the three concentrations of nitrous oxide. There were no changes in pure tone threshold with the various concentrations of nitrous oxide (table II). Thus while nitrous oxide TABLE I. Amplitude of conical evoked potentials to 70-dB HL and 40-dB HL tone bursts at 2000 Hz. t = 5.27; P < Mean amplitude (uv) Standard deviation 70dB HL db HL TABLE II. Changes in subjective auditory threshold at 2000 Hz with increasing nitrous oxide concentrations and subtracting this value from individual measurements. The mean corrected N1/P2 amplitudes were plotted for each subject for the different nitrous oxide concentrations (fig. 2). A significant negative correlation was found between Threshold change (db) Standard deviation Nitrous oxide concn. 10% 20% 40%

4 NITROUS OXIDE AND EVOKED POTENTIALS 609 TABLE III. Effects of nitrous oxide concentration on latency for 30-dB reduction in stimulus intensity or a the Nl and P2 potentials (mean (SD)). F ratios: Nl = 4.1, P2 = 3.5. *P < 0.05, **P < 0.02 Latency Nl P2 (ms) 100% O 2 10% 106 (9) 104 (6) 184(16) 186(19) Nitrous oxide concn 20% 40% 103 (7) 93 (12)* 180(18) : 172(7)** TABLE IV. Effects of stimulus intensity on the latencies of the cortical evoked potentials {mean (SD)) Latency Nl P2 (ms) 70 db HL 92(8) 173.8(16) 40dBHL 95(6) (9) P ns ns affected cortical auditory evoked potential threshold, no effect was observed on subjective pure tone threshold. A decrease in the latencies of both Nl and P2 was produced by increasing the concentration of nitrous oxide (table III). This was significant only when the nitrous oxide concentration reached 40 % (P < 0.02 for Nl; P < 0.05 for P2). Given that the amplitude change with increasing nitrous oxide concentration paralleled the change in amplitude produced by reduction in stimulus, a similar relationship might have been expected for the latency effect. The latencies of Nl and P2 of the cortical evoked potential were measured to 70 db HL and 40 db HL (table IV). No significant decrease in the latencies of either Nl or P2 was observed. Indeed, the mean latencies of the controls (Nl =93 ms, P2 = 172 ms) compared more favourably with the mean latencies at 40% nitrous oxide (Nl =93 ms, P2 = 172 ms) than with oxygen (Nl = 106 ms, P2 = 184 ms). Thus reduction in the stimulus did not have the same effect on latency as did increasing the nitrous oxide concentration. DISCUSSION The amplitude changes in the evoked potentials produced by nitrous oxide support the findings of Jarvis and Lader [3] and Fenwick and colleagues [4]. Accompanying these amplitude changes, there was a systematic increase in evoked potential threshold. The amplitude decrement produced by nitrous oxide was similar to that produced by a threshold increase of the same order. Thus the amplitude changes in cortical evoked potentials reported widely may be accounted for by the parallel changes in evoked potential threshold. There was no increase in evoked potential latencies with nitrous oxide; indeed latencies tended to decrease. This is evidence against latency jitter or desynchronization as a cause of the reduced amplitude produced by nitrous oxide. An alternative explanation for the latency shift is a latency increase produced by increasing concentration of oxygen. Although cortical evoked potential threshold was increased with increasing nitrous oxide concentration, there was no change in subjective pure tone threshold. That is, the cortical auditory evoked potential was not linked uniquely to subjective hearing levels. The neurophysiological pathways necessary for hearing as assessed through subjective pure tone tests do not appear to depend on the underlying neurophysiological substrate responsible for the cortical auditory evoked potentials. Our results support those of Fenwick and colleagues [4], who reported an increase in hearing sensitivity among several subjects breathing nitrous oxide. Our findings also support the view that the subjective acuity of hearing and the amplitude of the cortical evoked potential are unrelated. Thornton and colleagues [1, 2, 6] have shown stable, quantitative dose-related changes in the earlier middle latency evoked potentials, with both nitrous oxide and Althesin. It must be asked if the induced amplitude changes in these earlier potentials are attributable also to increased threshold for the evoked potentials. Irrespective of the mechanism whereby these changes are produced, auditory evoked potentials may be useful for assessing the depth of anaesthesia and the level of recovery. REFERENCES 1. Thornton C, Heneghan CPH, James MFM, Jones JG. Effects of halothane or enflurane with controlled ventilation on auditory evoked potentials. British Journal of Anaesthesia 1984; 56: Thornton C, Heneghan CPH, Navaratnarajah M, Bateman PE, Jones JG. Effect of etomidate on the auditory evoked response in man. British Journal of Anaesthesia 1985; 57: Jarvis MJ, Lader MH. The effects of nitrous oxide on the auditory evoked response in a reaction time task. Psychopharmacologia 1971; 20:

5 610 BRITISH JOURNAL OF ANAESTHESIA 4. Fenwick P, Bushman J, Howard R, Perry I, Gamble T. oxide upon sensory thresholds. Canadian Anaesthetists' Contingent negative variation and evoked potential am- Society Journal 1960; 7: plitude as a function of inspired nitrous oxide 6. Thornton C, Heneghan CPH, Navaratnarajah M, Jones concentration. Electroencephalography and Clinical JG. Selective effect of Althesin on the auditory evoked Neurophysiology 1979; 47: response in man. British Journal of Anaesthesia 1986; 58: 5. Bums DB, Robson JG, Welt PJL. The effect so nitrous 422^127.